Reasoning Validator

The reasoning validator prevents polished but invalid argument chains.

What it does

It evaluates premise-to-conclusion logic, identifies hidden assumptions, flags inference jumps, and marks claim chains as valid, weak, or invalid.

Output

A reasoning validation report with:

• claim-by-claim logic verdicts
• identified reasoning errors
• rewrite or evidence-gap fixes

Used by

Primary in /reasoning-validate, and as a quality gate in /iterate and final synthesis phases.